Exploring magnetocaloric materials by ab-initio methods

Abstract

This thesis explores the characterization of magnetocaloric materials from first-principles calculations, emphasizing entropy variation associated with the magnetocaloric effect. The study happens in the context of the search for new magnetocaloric materials to be applied in domestic magnetic refrigerators, as environmentally friendly and energy-efficient alternatives to conventional vapor-compression devices. The study involves benchmarking entropy calculations in systems like FeRh, which exhibits a first-order metamagnetic transition, and Gd, with a second-order ferromagnetic-paramagnetic transition. Different levels of approximations are examined and compared against experimental data, highlighting the need to distinguish between first-order and second-order transitions in the approach taken. The tests underscore the necessity of calculating vibrational and elastic properties or both phases to accurately calculate the entropy variation. This insight is applied in the study of Mn,.,Fe,.,NiSi,.,Al0.05, with results consistent with experimental data. Furthermore, the relationship between structural changes and magnetic properties is investigated, in particular for pressure-induced polymorphs in Gd and the phase transition in iSi,.,,AJ,.,. In the case of Gd, it was shown that variations in magnetic ordering temperature under pressure could be explained through a model based on the formation and accumulation of stacking faults. For the Mno.sl'eo.sNiSi,.,Al,.os system, the adoption of a magnetic composite model, in conjunction with experimental data, allowed to determine that the magnetostructural transition in these compounds is predominantly driven by the lattice subsystem. The results positively confirm the feasibility of using first-principles entropy estimates as an effective screening tool in high-throughput studies for magnetocaloric materials. A promising workflow is proposed, demonstrating potential in its initial results. Through comparison with experimental data, the derived routes offer valuable insights for the further refinement of workflow. This approach aims to enhance accuracy and systematically manage complex systems, highlighting a path forward for future advancements. Lastly, the introduction of a novel scaling scheme in Monte Carlo simulations enhancing accuracy across various temperatures, represents a potential advancement in the field of magnetic simulations.